CN105447187A - Webpage search method and system - Google Patents
Webpage search method and system Download PDFInfo
- Publication number
- CN105447187A CN105447187A CN201510945454.0A CN201510945454A CN105447187A CN 105447187 A CN105447187 A CN 105447187A CN 201510945454 A CN201510945454 A CN 201510945454A CN 105447187 A CN105447187 A CN 105447187A
- Authority
- CN
- China
- Prior art keywords
- query word
- real
- search
- default
- retrieval node
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
- G06F16/972—Access to data in other repository systems, e.g. legacy data or dynamic Web page generation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a webpage search method and system. According to the webpage search method and system, under the condition that real-time flow of the webpage search system is increased to an overload level and preset search results corresponding to target query words do not exist in a cache system, relatively few retrieval nodes relative to normal flow are selected to execute a real-time search task on the target query words, and the purpose of reducing loads of the retrieval nodes is achieved by reducing the number of real-time search results returned for the target query words, so that the response speed of the whole system is increased. In addition, when the real-time flow reaches a disaster-tolerant level, on one hand, the real-time search task are executed on the target query words through small amounts of the retrieval nodes to obtain the real-time search results, and on the other hand, the preset search results are obtained from the cache system to serve as a compensation, so that the loads of the retrieval nodes can be reduced and the completeness of the search results can be ensured.
Description
Technical field
The present invention relates to search engine optimization technical field, particularly relate to a kind of web search method and system.
Background technology
Along with the development of internet and universal, increasing user carries out Webpage search by the browser in various terminal device, to obtain the information that oneself needs.Fig. 1 shows the Organization Chart of web page search system; This web page search system comprises router (Router) 101, buffer memory (Cache) system 102 and retrieval node (Searcher) array 103, wherein, comprise in retrieval node (Searcher) array 103 M capable N row altogether M*N retrieve node, router one 01 maintains and the linking of each retrieval node all the time; In practical application, because Webpage search task amount is large, web page search system comprises the retrieval node array of multiple router and correspondence usually.
The principle of work of above-mentioned web page search system is as follows: when router one 01 receives the query word of client transmission, the Search Results whether having this query word corresponding is inquired about at caching system 102, if had, then the direct Search Results by this buffer memory is back to client; If caching system 102 does not have the Search Results that this query word of buffer memory is corresponding, then in each row of retrieval node array 103, selected at least one retrieval node (being namely total to selected at least N number of retrieval node) is searched for described query word, after the Search Results obtaining described retrieval node, on the one hand this Search Results is back to client, on the other hand the query word of this Search Results and correspondence is stored in described caching system, so that when receiving identical query word next time, from described caching system, directly obtain corresponding Search Results.
Based on above-mentioned principle of work, when the flow (namely at the appointed time search mission amount to be dealt with) of router is uprushed (such as, because optical cable damages, the fortuitous events such as power-off cause one or several router to quit work, other routers normally worked share the search mission of out-of-work router, the router traffic of normal work is then made to uprush), the search overhead of corresponding search node also can significantly increase, even exceed its upper loading limit, so high search overhead cannot be responded in time, cause search speed slack-off, whole web page search system stability reduces machine of even delaying.
Summary of the invention
For overcoming Problems existing in correlation technique, the invention provides a kind of web search method and system.
First aspect present invention, provides a kind of web search method, comprising:
The real-time traffic of monitoring web page search system;
When described real-time traffic is between default overload flow threshold and default disaster tolerance flow threshold, and, when there is not the default query word identical with target query word in the caching system of described web page search system, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission; Wherein, described default overload flow threshold is less than described default disaster tolerance flow threshold;
In the retrieval node array of described web page search system, select Q retrieval node, and trigger the real-time search task of Q retrieval node execution to described target query word selected, obtain real-time search result.
In conjunction with first aspect, in the first feasible embodiment of first aspect, described method also comprises:
Described real-time search result is stored in the buffer unit of corresponding retrieval node.
In conjunction with the first feasible embodiment of first aspect, in the embodiment that first aspect the second is feasible, described method also comprises:
The retrieval node number Q of described target query word and correspondence is stored in described caching system.
In conjunction with first aspect, or the first feasible embodiment of first aspect, or the embodiment that first aspect the second is feasible, in the third feasible embodiment of first aspect, the described ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission, comprising:
According to formula
calculate the retrieval node number Q of this search mission;
Wherein, W
newrepresent described real-time traffic, W represents that the normal discharge that described web page search system can be born, N represent the columns of described retrieval node array.
In conjunction with first aspect, or the first feasible embodiment of first aspect, or the embodiment that first aspect the second is feasible, in first aspect the 4th kind of feasible embodiment, Q retrieval node in the retrieval node array of the described web page search system of described selection, comprising:
Q retrieval node needed for this search mission is selected by tumbling-type back-and-forth method.
In conjunction with first aspect, in first aspect the 5th kind of feasible embodiment, described method also comprises:
When described real-time traffic is greater than described default disaster tolerance flow threshold, ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission, and in described caching system, search the default query word with target query word semi-match;
Select Q retrieval node in the retrieval node array of described web page search system, and trigger the real-time search task of Q retrieval node execution to described target query word chosen, obtain real-time search result;
Described real-time search result and preset search result corresponding to described default query word are merged, obtains the target search result that described target query word is corresponding.
In conjunction with first aspect the 5th kind of feasible embodiment, in first aspect the 6th kind of feasible embodiment, described method also comprises:
According to cryptographic hash and the word segmentation result thereof of each default query word stored in described caching system, in described caching system, build the inverted index of each participle, obtain cryptographic hash corresponding to each participle and fall row chain;
Describedly in described caching system, search the default query word with target query word semi-match, comprising:
Determine each target participle that target query word is corresponding and weighted value corresponding to each target participle;
By target participle described in weighted value select progressively from high to low, until the weighted value sum of the target participle chosen is not less than default weight threshold;
Judge whether that cryptographic hash that multiple target participles of choosing described in there is at least one common factor cryptographic hash is arranged in simultaneously are corresponding falls row chain;
If there is described common factor cryptographic hash, then default query word corresponding for described common factor cryptographic hash is labeled as the default query word with target query word semi-match.
In conjunction with first aspect the 6th kind of feasible embodiment, in first aspect the 7th kind of feasible embodiment, described method also comprises:
The retrieval node number Q of described target query word and correspondence, common factor cryptographic hash are stored in described caching system.
The second aspect of the embodiment of the present invention, provides a kind of web page search system, comprising: flow monitoring unit and overload processing unit;
Wherein, described flow monitoring unit is used for, the real-time traffic of monitoring web page search system;
Described overload processing unit is used for, process described real-time traffic between default overload flow threshold and default disaster tolerance flow threshold time search mission to target query word; Wherein, described default overload flow threshold is less than described default disaster tolerance flow threshold;
Described overload processing unit comprises: overload node calculate unit and overload sensor selection problem unit;
Wherein, described overload node calculate unit is used for, when described real-time traffic is between default overload flow threshold and default disaster tolerance flow threshold, and when there is not the default query word identical with target query word in the caching system of described web page search system, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission;
Described overload sensor selection problem unit is used for, and selects Q retrieval node in the retrieval node array of described web page search system, and triggers the real-time search task of Q retrieval node execution to described target query word chosen, and obtains real-time search result.
In conjunction with second aspect, in the first feasible embodiment of second aspect, described overload processing unit also comprises:
Overload result first buffer unit, for being stored in the buffer unit of corresponding retrieval node by described real-time search result.
In conjunction with the first feasible embodiment of second aspect, in the embodiment that second aspect the second is feasible, described overload processing unit also comprises: overload result second buffer unit, for being stored in described caching system by the retrieval node number Q of described target query word and correspondence.
In conjunction with second aspect, or the first feasible embodiment of second aspect, or the embodiment that second aspect the second is feasible, in the third feasible embodiment of second aspect, described overload node calculate unit is specifically configured to:
According to formula
calculate the retrieval node number Q of this search mission;
Wherein, W
newrepresent described real-time traffic, W represents that the normal discharge that described web page search system can be born, N represent the columns of described retrieval node array.
In conjunction with second aspect, or the first feasible embodiment of second aspect, or the embodiment that second aspect the second is feasible, in second aspect the 4th kind of feasible embodiment, described overload sensor selection problem unit is specifically configured to:
Q retrieval node needed for this search mission is selected by tumbling-type back-and-forth method.
In conjunction with second aspect, in second aspect the 5th kind of feasible embodiment, described system also comprises: disaster tolerance processing unit, the search mission to target query word when being greater than described default disaster tolerance flow threshold for the treatment of described real-time traffic;
Described disaster tolerance processing unit comprises: disaster tolerance node calculate unit, query word matching unit, disaster tolerance sensor selection problem unit and disaster tolerance result merge cells;
Disaster tolerance node calculate unit, for when described real-time traffic is greater than described default disaster tolerance flow threshold, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission;
Query word matching unit, for when described real-time traffic is greater than described default disaster tolerance flow threshold, searches the default query word with target query word semi-match in described caching system;
Disaster tolerance sensor selection problem unit, for selecting Q retrieval node in the retrieval node array of described web page search system, and triggering the real-time search task of Q retrieval node execution to described target query word chosen, obtaining real-time search result;
Disaster tolerance result merge cells, for described real-time search result and preset search result corresponding to described default query word being merged, obtains the target search result that described target query word is corresponding.
In conjunction with second aspect the 5th kind of feasible embodiment, in second aspect the 6th kind of feasible embodiment, described system also comprises:
Index construct unit, for according to the cryptographic hash of each default query word stored in described caching system and word segmentation result thereof, builds the inverted index of each participle in described caching system, obtains cryptographic hash corresponding to each participle and falls row chain;
Accordingly, described query word matching unit comprises:
Participle Weight Acquisition unit, for determining each target participle that target query word is corresponding and weighted value corresponding to each target participle;
Participle selection unit, for pressing target participle described in weighted value select progressively from high to low, until the weighted value sum of the target participle chosen is not less than default weight threshold;
Common factor judging unit, for judging whether that cryptographic hash that multiple target participles of choosing described in there is at least one common factor cryptographic hash is arranged in are corresponding falls row chain simultaneously, if there is described common factor cryptographic hash, then default query word corresponding for described common factor cryptographic hash is labeled as the default query word with target query word semi-match.
In conjunction with second aspect the 6th kind of feasible embodiment, in second aspect the 7th kind of feasible embodiment, described disaster tolerance processing unit also comprises:
Disaster tolerance result cache unit, for being stored in the retrieval node number Q of described target query word and correspondence, common factor cryptographic hash in described caching system.
From above technical scheme, the embodiment of the present application is when the real-time traffic of web page search system is increased to overload level and there is not the preset search result corresponding with target query word in caching system, under choosing relative normal discharge, the retrieval node of less number performs real-time search task to target query word, reaching the object of the load reducing each retrieval node by sacrificing the real-time search fruiting quantities returned for each target query word, improving the response speed of whole system.In addition, when real-time traffic reaches disaster tolerance rank, the present embodiment determines according to the ratio of real-time traffic and normal discharge the retrieval node number Q performing this search mission on the one hand, and in retrieval node array, select Q to retrieve node to target query word execution real-time search task, obtain real-time search result, on the other hand also by obtaining preset search result as compensation from caching system; Namely the Search Results being back to client under disaster tolerance rank is made up of two parts: the real-time search result that retrieval node obtains and the preset search result obtained from caching system, both can reduce the load of retrieval node, the integrality of Search Results can have been ensured again.
Should be understood that, it is only exemplary and explanatory that above general description and details hereinafter describe, and can not limit the disclosure.
Accompanying drawing explanation
Accompanying drawing to be herein merged in instructions and to form the part of this instructions, shows embodiment according to the invention, and is used from instructions one and explains principle of the present invention.
Fig. 1 is the Organization Chart of web page search system in correlation technique.
Fig. 2 is the process flow diagram of a kind of web search method according to an exemplary embodiment.
Fig. 3 is the process flow diagram of the another kind of web search method according to an exemplary embodiment.
Fig. 4 is the process flow diagram of another web search method according to an exemplary embodiment.
Fig. 5 is the process flow diagram of another web search method according to an exemplary embodiment.
Fig. 6 is the structured flowchart of a kind of web page search system according to an exemplary embodiment.
Fig. 7 is the structured flowchart of the another kind of web page search system according to an exemplary embodiment.
Fig. 8 is the structured flowchart of another web page search system according to an exemplary embodiment.
Embodiment
Here will be described exemplary embodiment in detail, its sample table shows in the accompanying drawings.When description below relates to accompanying drawing, unless otherwise indicated, the same numbers in different accompanying drawing represents same or analogous key element.Embodiment described in following exemplary embodiment does not represent all embodiments consistent with the present invention.On the contrary, they only with as in appended claims describe in detail, the example of apparatus and method that aspects more of the present invention are consistent.
Fig. 2 shows the process flow diagram of a kind of web search method that the embodiment of the present application provides.As shown in Figure 2, the method comprises the following steps.
The real-time traffic of S11, monitoring web page search system.
S12, when described real-time traffic is between default overload flow threshold and default disaster tolerance flow threshold, and, when there is not the default query word identical with target query word in the caching system of described web page search system, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission.
Wherein, described default overload flow threshold is less than described default disaster tolerance flow threshold; In conjunction with practical situations, described default overload flow threshold can be set as that this web page search system normally works the maximum flow (concrete numerical value is determined according to system performance) that can bear, described default overload flow threshold be set as this web page search system normally works 2 times of the maximum flow that can bear.
By above-mentioned default overload flow threshold and default disaster tolerance flow threshold, the flow of web page search system is divided into three interval/ranks, the risk that web page search system exists under different flow rank is different.Suppose the real-time traffic W of web page search system
new, presetting overload flow threshold is W
1, presetting disaster tolerance flow threshold is W
2if, then W
new< W
1, illustrate that real-time traffic is in default level, i.e. the normal discharge scope that can bear of system, under this default level, web page search system adopts each search mission of prior art process can meet application demand, if but W
1< W
new< W
2, illustrate that real-time traffic is in overload level, under this traffic class, if web page search system still adopts each search mission of prior art process, its processing speed can be slack-off, if W
new> W
2, illustrate that real-time traffic is in disaster tolerance rank, under this traffic class, if still adopt each search mission of prior art process, can there is machine risk of delaying in web page search system.Different disposal routes is taked for different ranks.
S13, in the retrieval node array of described web page search system, select Q retrieval node, and trigger the real-time search task of select Q retrieval node execution to described target query word, obtain real-time search result.
Above-mentioned real-time search result and target search result corresponding to target query word, returned the client that this target query word is corresponding, user just can see the Search Results wanted.
In addition, if there is the default query word identical with target query word in described caching system, then the preset search result that in described caching system, the described default query word of buffer memory is corresponding is directly returned.This situation and prior art disposal route similar, repeat no more herein.
Technique scheme describes real-time traffic and is in overload level (i.e. W
1< W
new< W
2) time the searching method that adopts of web page search system.System architecture shown in composition graphs 1, when real-time traffic is in overload level, similarly to the prior art, router one 01 is after receiving the target query word from client, first in caching system 102, whether identical query word has been stored, if no, then need to perform real-time search by the retrieval node in the retrieval node array 103 of M capable N row (M*N) retrieval node composition; In this case, if the retrieval node number that this real-time search is searched in the execution still chosen according to prior art is identical under normal circumstances with flow, then because flow system flow is uprushed, the search mission that average each retrieval node will perform in same time also can correspondingly increase, and namely retrieves the corresponding increase of load of node.
In view of this, the present embodiment determines according to the ratio of described real-time traffic and normal discharge the retrieval node number Q performing this search mission, ensure that Q is less than retrieval node number selected in normal discharge situation, thus reduce the average load of each retrieval node.
Such as, in supposing the system totally 10 retrieval nodes, in normal discharge situation, for each target query word, these 10 retrieval nodes all need to perform real-time search respectively, return 10 real-time search results, then according to prior art, when real-time traffic is in overload level, for x target query word, average each retrieval node needs to perform x real-time search, but according to the embodiment of the present application, reduce the retrieval node number Q each target query word being performed to real-time search, as got Q=5, then for x target query word, average each retrieval node only need perform x/2 real-time search, although the real-time search result number that each target query word returns is reduced to 5, but the load each can being retrieved node reduces half, namely the present embodiment is by sacrificing a part of real-time search result of each target query word, ensure response speed and the stability of web page search system.
Under overload level, flow system flow is uprushed, if be still stored in caching system according to prior art by the real-time search result of retrieval node, caching system is also difficult to bear so high load, easily causes caching system to collapse.Therefore, the embodiment of the present application is under overload level, real-time search task is stored in the buffer unit of corresponding retrieval node oneself, namely retrieve node 11 search for the real-time search result obtained be stored in retrieve node 11 buffer unit in, retrieval node 12 is searched for the real-time search result obtained and is stored in the buffer unit of retrieval node 12, and the rest may be inferred.Visible, above-mentionedly under overload level, real-time search result to be stored in the buffer unit of corresponding retrieval node, the storage burden of caching system can be alleviated, avoid caching system to collapse because of load too high.
From above technical scheme, the embodiment of the present application is when the real-time traffic of web page search system is increased to overload level and there is not the preset search result corresponding with target query word in caching system, under choosing relative normal discharge, the retrieval node of less number performs real-time search task to target query word, reach the object of the load reducing each retrieval node by sacrificing the real-time search fruiting quantities returned for each target query word, thus improve the response speed of whole system.
With reference to Fig. 3, in the embodiment that the application one is feasible, above-mentioned web search method can also comprise the steps:
S14, described real-time search result is stored in the buffer unit of corresponding retrieval node.
Real-time search result is stored in the buffer unit of corresponding retrieval node by the present embodiment under overload level, can when this retrieval node receive the real-time search task for identical target query word next time, from oneself buffer unit, directly take out the Search Results of institute's buffer memory; When namely retrieving node execution real-time search task in step s 13, first search in the buffer unit of oneself and whether store with target query word to corresponding real-time search result, if had, then directly return the real-time search result of institute's buffer memory, if no, then in internet carry out real-time search.Visible, above-mentionedly real-time search result is stored in the buffer unit of corresponding retrieval node under overload level, the load of retrieval node can be alleviated further, improve the response speed of whole system.
In the embodiment that the application one is feasible, above-mentioned steps S15 specifically can according to the retrieval node number Q performing real-time search under following formula determination overload level:
wherein, W
newrepresent described real-time traffic, W represents that the normal discharge that described web page search system can be born, N represent the columns of described retrieval node array.Because W and N is definite value, and under overload level, W
new> W, therefore according to above-mentioned formula, Q must be less than N, and Q is along with real-time traffic W
newincrease and reduce, thus avoid retrieve node overload.
In the embodiment that the application one is feasible, Q retrieval node in the retrieval node array of the described web page search system of the selection described in above-mentioned steps S16, comprising: select Q retrieval node needed for this search mission by tumbling-type back-and-forth method.
Namely above-mentioned tumbling-type back-and-forth method goes forward one by one the retrieval node that predetermined number starts to select needed for this on the basis of upper once selection result.Such as, suppose to have in web search system eight the retrieval nodes being numbered 1 to 8, Q=5 is calculated under overload level, then when for first aim query selection retrieval node, the real-time search task of retrieval node execution to first aim query word being numbered 1 to 5 can be selected, when for second target query selection retrieval node, 3 (i.e. the 8-5) that go forward one by one are individual, select from the retrieval node being numbered 4, namely the retrieval node being numbered 4 to 8 is finally selected, when for the 3rd target query word selection retrieval node, continue to go forward one by one 3, select from the retrieval node being numbered 7, namely final selecting is numbered 7, the retrieval node of 8 and 1 to 3.And for example, suppose that web page search system comprises 180 row retrieval nodes altogether, Q=120,60 row that then go forward one by one roll and select retrieval node, namely to first aim query word, from the often row the 1 to 120 row, select a retrieval node to perform search mission respectively, to second target query word, from the often row the 61 to 180 row, select a retrieval node to perform search mission respectively, to the 3rd target query word, from the often row the 121 to 180 row and the 1 to 60 row, select a retrieval node to perform search mission respectively, the rest may be inferred.
The present embodiment selects by tumbling-type back-and-forth method the retrieval node at every turn performing real-time search task, a part can be avoided to retrieve, and node searching number of times is too much, load too high, and another part retrieval node searching number of times is very few, load is lower, is about to repeatedly real-time search task mean allocation and retrieves node to each.
With reference to Fig. 4, in the embodiment that the application one is feasible, described default overload flow threshold (being namely in default level) is less than at real-time traffic, and, when there is not the default query word identical with target query word in the caching system of described web page search system, described web search method comprises the steps:
S22, be less than described default overload flow threshold when real-time traffic, and, from each retrieving node array (the capable N of M arranges) arranges, one is selected to retrieve node respectively when there is not the default query word identical with target query word in the caching system of described web page search system, trigger the selected real-time search task of N number of retrieval node execution to target query word, obtain real-time search result.
That is, under default level, retrieval node number Q=N, can ensure the integrality of Search Results.Concrete, can come to select one as the retrieval node in these row, target query word being performed to real-time search task from the M same row retrieval node according to the cryptographic hash of target query word (Hash, also known as hashed value).
In addition, under default level, if there is the default query word identical with this target query word in the caching system of described web page search system, then directly return the preset search result that this default query word is corresponding.
Further, the web search method under default level also comprises:
S23, the target query word of described real-time search result and correspondence to be stored in caching system.
Distributed storage mode can be adopted in caching system to carry out the Search Results of storing queries word and correspondence; Concrete, in above-mentioned steps S23, using target query word as major key (key), the real-time search result of correspondence can be stored in caching system as key assignments (value).Above-mentioned distributed storage mode has the features such as inquiry velocity is fast, support height is concurrent, can ensure the response speed of caching system.
Still with reference to Fig. 4, corresponding with above-mentioned steps S23, above-mentioned web search method can also comprise the steps: under overload level
S15, the target query word of described retrieval node number Q and correspondence to be stored in described caching system.
Corresponding to above-mentioned distributed storage mode, in the storing process of step S15, can using target query word as major key, corresponding retrieval node number Q is as key assignments.
Visible, different from step S23, in step S15, namely under overload level, only store target query word and corresponding Q value in caching system, and do not store real-time search result (according to step S14, under overload level, real-time search result is stored in the buffer unit of corresponding retrieval node), thus storage space shared in caching system can be greatly reduced, under avoiding high flow capacity, caching system collapses because memory load is too high.When receiving identical query word next time, the Q value that this query word is corresponding can be found in caching system, and then pass through hash algorithm, Q retrieval node of the real-time search task performed last time this query word can be learnt, thus Search Results corresponding to this query word can be read in the buffer unit of this Q retrieval node; Certainly, also can according to the actual requirements and the factor such as real-time flow data, determine whether still to adopt this Q to retrieve node, the need of several retrieval nodes of increase, to ensure the integrality of Search Results.Relative to recalculating Q value and again performing real-time search task by retrieval node, the present embodiment realizes the L2 cache to Search Results jointly by above-mentioned steps S14 and S15, can under the prerequisite alleviating caching system load, improve search efficiency and the response speed of system under overload level.
With reference to Fig. 5, in the embodiment that the application one is feasible, based on embodiment illustrated in fig. 2, when real-time traffic is greater than described default disaster tolerance flow threshold (being namely in disaster tolerance rank), described web search method comprises the steps:
S32, when described real-time traffic is greater than described default disaster tolerance flow threshold, ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission, and in described caching system, search the default query word with target query word semi-match;
S33, Q the retrieval node selected in the retrieval node array of described web page search system, and trigger the real-time search task of Q retrieval node execution to described target query word chosen, obtain real-time search result;
S34, described real-time search result and preset search result corresponding to described default query word to be merged, obtain the target search result that described target query word is corresponding.
When reaching disaster tolerance rank, mean that flow system flow has reached 3 times even 10 times that system can bear flow, system probably cannot bear so high load so that machine of delaying, and cannot provide search service.For this situation, the present embodiment is when receiving target query word, the retrieval node number Q performing this search mission is determined on the one hand according to the ratio of real-time traffic and normal discharge, and in retrieval node array, select Q to retrieve node to target query word execution real-time search task, obtain real-time search result.Because under disaster tolerance rank, real-time traffic is too high, therefore the Q value finally determined can be very little, so that the real-time search result that retrieval node returns also can be little.Therefore, the present embodiment on the other hand also by obtaining preset search result as compensation from caching system, namely the Search Results being back to client under disaster tolerance rank is made up of two parts: the real-time search result that retrieval node obtains and the preset search result obtained from caching system, both retrieval node load can have been avoided too high, the integrality of Search Results can have been ensured again.
Concrete, consider that in caching system, existence is very little with the probability of the identical default query word of target query word, the present embodiment adopts the mode of semi-match to search described default query word, namely in caching system, search the default query word with target query word semi-match, and then read and return preset search result corresponding to these similar default query words.
In the embodiment that the application one is feasible, in caching system, the semi-match formula of target query word is searched under disaster tolerance rank can being realized by inverted index.Wherein, the embodiment of the present application also comprises the steps, to build above-mentioned inverted index in caching system:
According to cryptographic hash and the word segmentation result thereof of each default query word stored in described caching system, in described caching system, build the inverted index of each participle, obtain cryptographic hash corresponding to each participle and fall row chain.
In the present embodiment, along with the increase of the Search Results of target query word new in caching system and correspondence, perform above-mentioned steps in real time and realize structure to described inverted index and renewal.Such as, cryptographic hash and the word segmentation result thereof of 3 query words in input-buffer system are as shown in the table.
| Query word | Cryptographic hash | Word segmentation result |
| Query1 | hash1 | A、B |
| Query2 | Hash2 | A、C、E、D |
| Query3 | Hash3 | B、C |
Upper table is considered as forward index, namely searches corresponding participle (Term) according to cryptographic hash, is converted to inverted index form, and row chain that what each participle was corresponding fall is as follows:
A→hash1hash2；
B→hash1hash3；
C→hash2hash3；
D→hash2；
E→hash2。
When there being new query word input-buffer system, directly upgraded by the mode inserted or increase newly on the basis of established inverted index, such as, the cryptographic hash of the query word Query4 of input-buffer system is Hash4, word segmentation result is " C, D, F ", then upgrade on the basis of the inverted index of above-mentioned A to E five formations of falling row chain, result is as follows:
A→hash1hash2；
B→hash1hash3；
C→hash2hash3hash4；
D→hash2hash4；
E→hash2；
F→hash4。
By above-mentioned inverted index, can search and relevant cryptographic hash according to participle, and then the query word can determining to be correlated with (the corresponding cryptographic hash of each query word).
Based on above-mentioned inverted index, described in above-mentioned steps S32, in caching system, search the default query word with target query word semi-match, specifically can comprise the steps:
Determine each target participle that target query word is corresponding and weighted value corresponding to each target participle;
By target participle described in weighted value select progressively from high to low, until the weighted value sum of the target participle chosen is not less than default weight threshold;
Judge whether that cryptographic hash that multiple target participles of choosing described in there is at least one common factor cryptographic hash is arranged in simultaneously are corresponding falls row chain;
If there is described common factor cryptographic hash, then default query word corresponding for described common factor cryptographic hash is labeled as the default query word with target query word semi-match.
Such as, carry out participle can obtain A, C, D tri-target participles to certain target query word, corresponding weighted value is respectively W
a, W
cand W
d, and, W
a>W
c>W
d; Then press weighted value order from high to low to add up, W
a+ W
cresult reach and preset inquiry threshold value W ' (i.e. W
a<W ' and W
a+ W
c>=W '), then selected target participle is A and C; Obtain row chain " A → hash1hash2 " and " C → hash2hash3hash4 " that A and C is corresponding, and to two fall row chain carry out intersection operation, determine existence common factor cryptographic hash hash2, then the default query word Query2 that common factor cryptographic hash hash2 is corresponding and target query word semi-match; And then in caching system, read preset search result corresponding to common factor cryptographic hash hash2 (namely presetting query word Query2), the real-time search result obtained with retrieval node, in the lump as the final Search Results of described target query word, is back to corresponding client.
Certainly, in particular cases, if there is no described common factor cryptographic hash, then illustrate in caching system there is not the default query word with target query word semi-match, similar preset search result can not be obtained from caching system, be finally back in the Search Results of client a small amount of real-time search result also just only comprising retrieval node and obtain.
From above step, the present embodiment realizes searching the semi-match formula of target query word in caching system by inverted index, simple, thus easily can get the relevant preset search result of target query word from caching system.
Corresponding to step S15 mentioned above and step S23, the Webpage searching result caching method that the present embodiment provides also comprises the steps: under disaster tolerance rank
The retrieval node number Q of described target query word and correspondence, common factor cryptographic hash are stored in described caching system.
Based on above-mentioned steps, in caching system, whole Search Results is not stored under disaster tolerance rank, and only store retrieval node number Q, the common factor cryptographic hash of target query word and correspondence, the memory load of caching system can be reduced on the one hand, on the other hand can also when system receives same target query word again, directly in caching system, find corresponding Q value and common factor cryptographic hash, and then real-time search result corresponding to quick obtaining and preset search result, improve the response speed of web page search system.
From above embodiment of the method, the real-time traffic of web page search system is divided into three ranks by the application, is followed successively by from low to high: default level, overload level and disaster tolerance rank; And provide different web search method for different ranks, ensure the system stability under each rank and response speed, avoid caching system or retrieval node overload, avoid system to delay machine.
In addition, the embodiment of the present application additionally provides a kind of computer-readable storage medium, such as, can be ROM, random access memory (RAM), CD-ROM, tape, floppy disk and optical data storage devices etc.; Have program stored therein in this computer-readable storage medium, when processor corresponding during the program in described storage medium is by web page search system performs, make web page search system can perform the part or all of step of the web search method recorded in said method embodiment.
The structured flowchart of a kind of web page search system that Fig. 6 provides for the embodiment of the present application; This system is made up of multiple performance element, and each performance element is divided in the equipment such as router, caching system, retrieval node.With reference to Fig. 6, the performance element forming this web page search system at least comprises: flow monitoring unit 100 and overload processing unit 200.
Wherein, this flow monitoring unit 100 for, the real-time traffic of monitoring web page search system, and determines the interval residing for described real-time traffic according to default overload flow threshold and default disaster tolerance flow threshold.Wherein, described default overload flow threshold is less than described default disaster tolerance flow threshold.Relative to system shown in Figure 1 framework, this flow monitoring unit 100 can be arranged in router one 01.
This overload processing unit 200 for, process described real-time traffic between default overload flow threshold and default disaster tolerance flow threshold time search mission to target query word.
Concrete, this overload processing unit 200 comprises: overload node calculate unit 201 and overload sensor selection problem unit 202.Relative to system shown in Figure 1 framework, overload node calculate unit 201 and overload sensor selection problem unit 202 can be arranged in router one 01,
This overload node calculate unit 201 for, when described real-time traffic is between default overload flow threshold and default disaster tolerance flow threshold, and when there is not the default query word identical with target query word in the caching system of described web page search system, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission;
This overload sensor selection problem unit 202 for, select in the retrieval node array of described web page search system Q retrieval node, and trigger the real-time search task of Q retrieval node execution to described target query word chosen, obtain real-time search result.
From above system architecture, the embodiment of the present application the embodiment of the present application is when the real-time traffic of web page search system is increased to overload level and there is not the preset search result corresponding with target query word in caching system, under choosing relative normal discharge, the retrieval node of less number performs real-time search task to target query word, reach the object of the load reducing each retrieval node by sacrificing the real-time search fruiting quantities returned for each target query word, and then improve the response speed of whole system.
In the embodiment that the application one is feasible, above-mentioned overload node calculate unit 201 is specifically configured to: according to formula
calculate the retrieval node number Q of this search mission; Wherein, W
newrepresent described real-time traffic, W represents that the normal discharge that described web page search system can be born, N represent the columns of described retrieval node array.
In another feasible embodiment of the application, above-mentioned overload sensor selection problem unit 202 is specifically configured to: select Q retrieval node needed for this search mission by tumbling-type back-and-forth method.
With reference to Fig. 7, in the embodiment that the application one is feasible, above-mentioned web page search system also comprises: default treatment unit 300, the search mission to target query word when being less than described default overload flow threshold (i.e. default level) for the treatment of real-time traffic.This default treatment unit 300 can comprise: default node selection unit 301 and acquiescence result cache unit 302.
This default node selection unit 301 can be arranged in router one 01, for being less than described default overload flow threshold when real-time traffic, and, from each retrieving node array (the capable N of M arranges) arranges, one is selected to retrieve node respectively when there is not the default query word identical with target query word in the caching system of described web page search system, trigger the selected real-time search task of N number of retrieval node execution to target query word, obtain real-time search result.
This acquiescence result cache unit 302 is arranged in caching system 102, for storing the target query word of described real-time search result and correspondence.
Corresponding to above-mentioned acquiescence result cache unit 302, above-mentioned overload processing unit 200 can also comprise: overload result first buffer unit 203 and overload result second buffer unit 204.
Wherein, overload result first buffer unit 203 is arranged in each retrieval node, for storing the described real-time search result that corresponding retrieval node obtains.
Overload result second buffer unit 204 is arranged in caching system 102, for the retrieval node number Q of the target query word that receives under storing overload level and correspondence.
Visible, under overload level, corresponding real-time search result is stored by overload result first buffer unit 203 be arranged in each retrieval node, overload result second buffer unit 204 in caching system stores target query word and corresponding Q value, thus storage space shared in caching system can be greatly reduced, under avoiding high flow capacity, caching system collapses because memory load is too high.When receiving identical query word next time, Q value corresponding to this query word can be found in overload result second buffer unit 204, and then pass through hash algorithm, Q retrieval node of the real-time search task performed last time this query word can be learnt, thus Search Results corresponding to this query word can be read in the buffer unit of this Q retrieval node.Relative to recalculating Q value and again performing real-time search task by retrieval node, the L2 cache that the present embodiment realizes Search Results by being arranged at overload result first buffer unit in retrieval node and overload result second buffer unit be arranged in caching system jointly, can under the prerequisite alleviating caching system load, improve search efficiency and the response speed of system under overload level.
With reference to Fig. 8, in the embodiment that the application one is feasible, above-mentioned web page search system also comprises: disaster tolerance processing unit 400, the search mission to target query word when being greater than described default disaster tolerance flow threshold (i.e. disaster tolerance rank) for the treatment of real-time traffic.This disaster tolerance processing unit 400 can comprise: disaster tolerance node calculate unit 401, query word matching unit 402, disaster tolerance sensor selection problem unit 403 and disaster tolerance result merge cells 404.
This disaster tolerance node calculate unit 401 is arranged in router one 01, and for when described real-time traffic is greater than described default disaster tolerance flow threshold, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission;
This query word matching unit 402 is arranged in caching system 102, for when described real-time traffic is greater than described default disaster tolerance flow threshold, searches the default query word with target query word semi-match in described caching system;
This disaster tolerance sensor selection problem unit 403 is arranged in router one 01, for selecting Q retrieval node in the retrieval node array of described web page search system, and trigger the real-time search task of Q retrieval node execution to described target query word chosen, obtain real-time search result;
This disaster tolerance result merge cells 404 is arranged in router one 01, for described real-time search result and preset search result corresponding to described default query word being merged, obtains the target search result that described target query word is corresponding.
From above result, when real-time traffic reaches disaster tolerance rank, the present embodiment determines according to the ratio of real-time traffic and normal discharge the retrieval node number Q performing this search mission on the one hand, and in retrieval node array, select Q to retrieve node to target query word execution real-time search task, obtain real-time search result, on the other hand also by obtaining preset search result as compensation from caching system; Namely the Search Results being back to client under disaster tolerance rank is made up of two parts: the real-time search result that retrieval node obtains and the preset search result obtained from caching system, both can reduce the load of retrieval node, the integrality of Search Results can have been ensured again.
In the embodiment that the application one is feasible, above-mentioned web page search system can also comprise: index construct unit, be arranged in caching system 102, for according to the cryptographic hash of each default query word stored in described caching system and word segmentation result thereof, in described caching system, build the inverted index of each participle, obtain cryptographic hash corresponding to each participle and fall row chain.Accordingly, above-mentioned query word matching unit 402 specifically can comprise:
Participle Weight Acquisition unit, for determining each target participle that target query word is corresponding and weighted value corresponding to each target participle;
Participle selection unit, for pressing target participle described in weighted value select progressively from high to low, until the weighted value sum of the target participle chosen is not less than default weight threshold;
Common factor judging unit, for judging whether that cryptographic hash that multiple target participles of choosing described in there is at least one common factor cryptographic hash is arranged in are corresponding falls row chain simultaneously, if there is described common factor cryptographic hash, then default query word corresponding for described common factor cryptographic hash is labeled as the default query word with target query word semi-match.
In the embodiment that the application one is feasible, corresponding to above-mentioned acquiescence result cache unit 302, overload result second buffer unit 203 and overload result second buffer unit 204, above-mentioned disaster tolerance processing unit 400 can also comprise: disaster tolerance result cache unit; This disaster tolerance result cache unit is arranged in caching system 102, during for being in disaster tolerance rank at flow system flow, stores retrieval node number Q, the common factor cryptographic hash of described target query word and correspondence.
Visible, in caching system, whole Search Results is not stored under disaster tolerance rank, and retrieval node number Q, the common factor cryptographic hash of target query word and correspondence is only stored by disaster tolerance result cache unit, the memory load of caching system can be reduced on the one hand, on the other hand can also when system receives same target query word again, directly in caching system, find corresponding Q value and common factor cryptographic hash, and then real-time search result corresponding to quick obtaining and preset search result, improve the response speed of web page search system.
Those skilled in the art, at consideration instructions and after putting into practice invention disclosed herein, will easily expect other embodiment of the present invention.The application is intended to contain any modification of the present invention, purposes or adaptations, and these modification, purposes or adaptations are followed general principle of the present invention and comprised the undocumented common practise in the art of the disclosure or conventional techniques means.Instructions and embodiment are only regarded as exemplary, and true scope of the present invention and spirit are pointed out by claim below.
Should be understood that, the present invention is not limited to precision architecture described above and illustrated in the accompanying drawings, and can carry out various amendment and change not departing from its scope.Scope of the present invention is only limited by appended claim.
Claims (16)
1. a web search method, is characterized in that, comprising:
The real-time traffic of monitoring web page search system;
When described real-time traffic is between default overload flow threshold and default disaster tolerance flow threshold, and, when there is not the default query word identical with target query word in the caching system of described web page search system, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission; Wherein, described default overload flow threshold is less than described default disaster tolerance flow threshold;
In the retrieval node array of described web page search system, select Q retrieval node, and trigger the real-time search task of Q retrieval node execution to described target query word selected, obtain real-time search result.
2. method according to claim 1, is characterized in that, also comprises:
Described real-time search result is stored in the buffer unit of corresponding retrieval node.
3. method according to claim 2, is characterized in that, also comprises:
The retrieval node number Q of described target query word and correspondence is stored in described caching system.
4. the method according to any one of claims 1 to 3, is characterized in that, the described ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission, comprising:
According to formula
calculate the retrieval node number Q of this search mission;
Wherein, W
newrepresent described real-time traffic, W represents that the normal discharge that described web page search system can be born, N represent the columns of described retrieval node array.
5. the method according to any one of claims 1 to 3, is characterized in that, Q retrieval node in the retrieval node array of the described web page search system of described selection, comprising:
Q retrieval node needed for this search mission is selected by tumbling-type back-and-forth method.
6. method according to claim 1, is characterized in that, also comprises:
When described real-time traffic is greater than described default disaster tolerance flow threshold, ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission, and in described caching system, search the default query word with target query word semi-match;
Select Q retrieval node in the retrieval node array of described web page search system, and trigger the real-time search task of Q retrieval node execution to described target query word chosen, obtain real-time search result;
Described real-time search result and preset search result corresponding to described default query word are merged, obtains the target search result that described target query word is corresponding.
7. method according to claim 6, is characterized in that, also comprises:
According to cryptographic hash and the word segmentation result thereof of each default query word stored in described caching system, in described caching system, build the inverted index of each participle, obtain cryptographic hash corresponding to each participle and fall row chain;
Describedly in described caching system, search the default query word with target query word semi-match, comprising:
Determine each target participle that target query word is corresponding and weighted value corresponding to each target participle;
By target participle described in weighted value select progressively from high to low, until the weighted value sum of the target participle chosen is not less than default weight threshold;
Judge whether that cryptographic hash that multiple target participles of choosing described in there is at least one common factor cryptographic hash is arranged in simultaneously are corresponding falls row chain;
If there is described common factor cryptographic hash, then default query word corresponding for described common factor cryptographic hash is labeled as the default query word with target query word semi-match.
8. method according to claim 7, is characterized in that, also comprises:
The retrieval node number Q of described target query word and correspondence, common factor cryptographic hash are stored in described caching system.
9. a web page search system, is characterized in that, comprising: flow monitoring unit and overload processing unit;
Wherein, described flow monitoring unit is used for, the real-time traffic of monitoring web page search system;
Described overload processing unit is used for, process described real-time traffic between default overload flow threshold and default disaster tolerance flow threshold time search mission to target query word; Wherein, described default overload flow threshold is less than described default disaster tolerance flow threshold;
Described overload processing unit comprises: overload node calculate unit and overload sensor selection problem unit;
Wherein, described overload node calculate unit is used for, when described real-time traffic is between default overload flow threshold and default disaster tolerance flow threshold, and when there is not the default query word identical with target query word in the caching system of described web page search system, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission;
Described overload sensor selection problem unit is used for, and selects Q retrieval node in the retrieval node array of described web page search system, and triggers the real-time search task of Q retrieval node execution to described target query word chosen, and obtains real-time search result.
10. system according to claim 9, is characterized in that, described overload processing unit also comprises: overload result first buffer unit, is specifically configured to:
Described real-time search result is stored in the buffer unit of corresponding retrieval node.
11. systems according to claim 10, is characterized in that, described overload processing unit also comprises:
Overload result second buffer unit, for being stored in the retrieval node number Q of described target query word and correspondence in described caching system.
12. systems according to any one of claim 9 to 11, it is characterized in that, described overload node calculate unit is specifically configured to:
According to formula
calculate the retrieval node number Q of this search mission;
Wherein, W
newrepresent described real-time traffic, W represents that the normal discharge that described web page search system can be born, N represent the columns of described retrieval node array.
13. systems according to any one of claim 9 to 11, it is characterized in that, described overload sensor selection problem unit is specifically configured to:
Q retrieval node needed for this search mission is selected by tumbling-type back-and-forth method.
14. systems according to claim 9, is characterized in that, also comprise: disaster tolerance processing unit, the search mission to target query word when being greater than described default disaster tolerance flow threshold for the treatment of described real-time traffic;
Described disaster tolerance processing unit comprises: disaster tolerance node calculate unit, query word matching unit, disaster tolerance sensor selection problem unit and disaster tolerance result merge cells;
Disaster tolerance node calculate unit, for when described real-time traffic is greater than described default disaster tolerance flow threshold, the ratio according to described real-time traffic and normal discharge determines the retrieval node number Q performing this search mission;
Query word matching unit, for when described real-time traffic is greater than described default disaster tolerance flow threshold, searches the default query word with target query word semi-match in described caching system;
Disaster tolerance sensor selection problem unit, for selecting Q retrieval node in the retrieval node array of described web page search system, and triggering the real-time search task of Q retrieval node execution to described target query word chosen, obtaining real-time search result;
Disaster tolerance result merge cells, for described real-time search result and preset search result corresponding to described default query word being merged, obtains the target search result that described target query word is corresponding.
15. systems according to claim 14, is characterized in that, also comprise:
Index construct unit, for according to the cryptographic hash of each default query word stored in described caching system and word segmentation result thereof, builds the inverted index of each participle in described caching system, obtains cryptographic hash corresponding to each participle and falls row chain;
Accordingly, described query word matching unit comprises:
Participle Weight Acquisition unit, for determining each target participle that target query word is corresponding and weighted value corresponding to each target participle;
Participle selection unit, for pressing target participle described in weighted value select progressively from high to low, until the weighted value sum of the target participle chosen is not less than default weight threshold;
Common factor judging unit, for judging whether that cryptographic hash that multiple target participles of choosing described in there is at least one common factor cryptographic hash is arranged in are corresponding falls row chain simultaneously, if there is described common factor cryptographic hash, then default query word corresponding for described common factor cryptographic hash is labeled as the default query word with target query word semi-match.
16. systems according to claim 15, is characterized in that, described disaster tolerance processing unit also comprises:
Disaster tolerance result cache unit, for being stored in the retrieval node number Q of described target query word and correspondence, common factor cryptographic hash in described caching system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510945454.0A CN105447187B (en) | 2015-12-15 | 2015-12-15 | Web search method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510945454.0A CN105447187B (en) | 2015-12-15 | 2015-12-15 | Web search method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105447187A true CN105447187A (en) | 2016-03-30 |
| CN105447187B CN105447187B (en) | 2017-09-22 |
Family
ID=55557363
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510945454.0A Active CN105447187B (en) | 2015-12-15 | 2015-12-15 | Web search method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105447187B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108470043A (en) * | 2018-02-27 | 2018-08-31 | 阿里巴巴集团控股有限公司 | A kind of acquisition methods and device of business result |
| CN108846094A (en) * | 2018-06-15 | 2018-11-20 | 江苏中威科技软件系统有限公司 | A method of based on index in classification interaction |
| CN110309390A (en) * | 2018-03-15 | 2019-10-08 | 广东神马搜索科技有限公司 | Index column indention method, apparatus and server suitable for search |
| CN113032436A (en) * | 2021-04-16 | 2021-06-25 | 苏州臻璇数据信息技术有限公司 | Searching method and device based on article content and title |
| CN114218013A (en) * | 2021-12-13 | 2022-03-22 | 北京字节跳动网络技术有限公司 | Searching method, searching device and electronic equipment storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103339597A (en) * | 2010-10-30 | 2013-10-02 | 布雷克公司 | Transforming search engine queries |
| US20140007171A1 (en) * | 2012-06-27 | 2014-01-02 | Nhn Corporation | Apparatus, method and computer readable recording medium for providing music related information by recognition of music output through television |
| CN103812949A (en) * | 2014-03-06 | 2014-05-21 | 中国科学院信息工程研究所 | Task scheduling and resource allocation method and system for real-time cloud platform |
| US20140236917A1 (en) * | 2013-02-19 | 2014-08-21 | Luxian Limited | Processor engine, integrated circuit and method therefor |
-
2015
- 2015-12-15 CN CN201510945454.0A patent/CN105447187B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103339597A (en) * | 2010-10-30 | 2013-10-02 | 布雷克公司 | Transforming search engine queries |
| US20140007171A1 (en) * | 2012-06-27 | 2014-01-02 | Nhn Corporation | Apparatus, method and computer readable recording medium for providing music related information by recognition of music output through television |
| US20140236917A1 (en) * | 2013-02-19 | 2014-08-21 | Luxian Limited | Processor engine, integrated circuit and method therefor |
| CN103812949A (en) * | 2014-03-06 | 2014-05-21 | 中国科学院信息工程研究所 | Task scheduling and resource allocation method and system for real-time cloud platform |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108470043A (en) * | 2018-02-27 | 2018-08-31 | 阿里巴巴集团控股有限公司 | A kind of acquisition methods and device of business result |
| CN110309390A (en) * | 2018-03-15 | 2019-10-08 | 广东神马搜索科技有限公司 | Index column indention method, apparatus and server suitable for search |
| CN108846094A (en) * | 2018-06-15 | 2018-11-20 | 江苏中威科技软件系统有限公司 | A method of based on index in classification interaction |
| CN113032436A (en) * | 2021-04-16 | 2021-06-25 | 苏州臻璇数据信息技术有限公司 | Searching method and device based on article content and title |
| CN114218013A (en) * | 2021-12-13 | 2022-03-22 | 北京字节跳动网络技术有限公司 | Searching method, searching device and electronic equipment storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105447187B (en) | 2017-09-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105447187A (en) | Webpage search method and system | |
| CN104679778B (en) | A kind of generation method and device of search result | |
| JP5466210B2 (en) | Table search device, table search method, and table search system | |
| RU2645266C1 (en) | Method and device for planning web-crowlers in accordance with keyword search | |
| CN104424199A (en) | Search method and device | |
| JP5014399B2 (en) | Search data management device | |
| CN103051543B (en) | A kind of process of route prefix, search, increase and delet method | |
| CN107015985B (en) | Data storage and acquisition method and device | |
| CN104809135B (en) | A kind of data access method and device | |
| CN104504003A (en) | Graph data searching method and device | |
| CN102761627A (en) | Cloud website recommending method and system based on terminal access statistics as well as related equipment | |
| CN102479207A (en) | Information search method, system and device | |
| CN105488201A (en) | Log inquiry method and system | |
| CN104809129A (en) | Method, device and system for storing distributed data | |
| CN104702625A (en) | Method and device for scheduling access request in CDN (Content Delivery Network) | |
| CN102915382A (en) | Method and device for carrying out data query on database based on indexes | |
| CN102035727A (en) | Method and equipment for processing routing | |
| CN106326475A (en) | High-efficiency static hash table implement method and system | |
| CN105721538A (en) | Data access method and apparatus | |
| JPH0619895A (en) | Method for storage of document processing information regarding item from limited text source | |
| CN105404675A (en) | Ranked reverse nearest neighbor space keyword query method and apparatus | |
| CN107798106A (en) | A kind of URL De-weight methods in distributed reptile system | |
| CN103955533A (en) | Device and method for acquiring page tree data based on buffering queue | |
| EP2840499A1 (en) | Parallel computer system, control program for management device, and method of controlling parallel computer system | |
| US7788284B2 (en) | System and method for knowledge based search system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200527 Address after: 310051 room 508, floor 5, building 4, No. 699, Wangshang Road, Changhe street, Binjiang District, Hangzhou City, Zhejiang Province Patentee after: Alibaba (China) Co.,Ltd. Address before: 510627 Guangdong city of Guangzhou province Whampoa Tianhe District Road No. 163 Xiping Yun Lu Yun Ping square B radio tower 12 layer self unit 01 Patentee before: GUANGZHOU SHENMA MOBILE INFORMATION TECHNOLOGY Co.,Ltd. |
|
| TR01 | Transfer of patent right |